Landscape features on the detection and probability of occupation of the black lion tamarin (Leontopithecus chrysopygus) in forest remnants of the upper and middle Paranapanema

• Grant number: 24/20380-7
• Support Opportunities: Scholarships in Brazil - Post-Doctoral
• Start date: February 01, 2025
• End date: January 31, 2027
• Field of knowledge: Biological Sciences - Ecology - Applied Ecology
• Principal Investigator: Laurence Marianne Vincianne Culot
• Grantee: André Luis Regolin
• Host Institution: Instituto de Biociências (IB). Universidade Estadual Paulista (UNESP). Campus de Rio Claro. Rio Claro , SP, Brazil
• Associated research grant: 21/06668-0 - Primate resilience in an anthropogenic landscape, AP.BTA.JP2

Abstract

Habitat loss, fragmentation, and degradation constitute the main threats to primate persistence globally. Therefore, it is essential to understand the relationships between species distribution and the spatial configuration of remaining habitats to inform the development of conservation and management programs. Habitat modeling is especially urgent for endangered species occurring in biodiversity hotspots, such as the black lion tamarin (Leontopithecus chrysopygus) in the Atlantic Forest. This study aims to identify how local variables and landscape structure variables influence the probability of occupancy and detection of the black lion tamarin in fragmented landscapes.It is expected that the probability of occupancy will be higher in well-preserved landscapes, where forest cover, ecological connectivity, and habitat quality are greater. Detection probability is anticipated to primarily vary with occasion-specific variables, such as wind intensity and direction, as well as structural variables, such as vegetation complexity and forest cover in the landscape. Twenty landscapes representing a gradient of disturbance levels will be studied. Landscape structure will be analyzed at multiple scales using a patch-landscape approach based on high-resolution thematic maps from the Mapbiomas project. Patch metrics (i.e., area, shape, and age of patches) and landscape metrics (i.e., functional connectivity, visibility, and compositional heterogeneity) will be calculated. Vegetation structural complexity near the sampling point, as well as wind intensity and direction, will describe local conditions.Species detection will be carried out using passive acoustic recorders, which will remain in the field for at least 12 days, operating during the species' peak activity periods (6:00-11:00 and 14:00-17:30). Two recorders will be installed in each landscape. Recordings will be analyzed with Raven Pro software, using sound pattern detectors through cross-correlations. The relationship between predictor variables and species records will be evaluated using occupancy models that account for imperfect species detection. Specifically, a multi-scale model will be adopted, which has three parameters: (1) ¿, the probability that a species occurs in the landscape (n=20); (2) ¿, the probability that the species is present at the sampling point (near the recorder), assuming the landscape is occupied by the species; and (3) p, the probability that the species is detected at the sampling point, given that it occurs in the landscape and is present at the sampling point. Several competing models will be fitted, and inference on the best models will be made using a model selection approach based on information criteria.